README.md

metadata

title: Palmoil
emoji: 📚
colorFrom: pink
colorTo: indigo
sdk: gradio
sdk_version: 5.20.1
app_file: app.py
pinned: false

Palm Oil Price Predictor

A machine learning application to predict palm oil prices using historical data.

Features

• Predict future palm oil prices based on a linear regression model
• View model performance metrics (MAE, MAPE, RMSE)
• Visualize historical data and predictions
• Retrain the model with updated data

Tech Stack

• Python
• Gradio (UI)
• Scikit-learn (ML)
• Pandas & NumPy (Data Processing)
• Hugging Face Spaces (Deployment)

Setup

1. Ensure palm_oil.xlsx contains your data with 'date' and 'nilai' columns
2. Run train_model.py locally to generate static/model/model.pkl
3. Deploy to Hugging Face Spaces with Gradio SDK

Usage

• Predict: From January 2025
• Model Insights: View performance metrics and historical data
• Retrain: Update the model with the latest data

Limitations

• Model retraining is temporary (resets on Space restart)
• Assumes linear relationship in data

Developed by [Ihsan Andrinal] | [https://www.linkedin.com/in/ihsan-andrinal/]